BASUG Quarterly Meeting Announcement

 

 

We are thrilled with the response we got to our Call-for-Papers for our Q4 Coders’ Corner meeting. This meeting will feature many presenters – some first-timers, some seasoned – covering a variety of topics. Please join us to learn lots of tips ‘n techniques from your fellow BASUG colleagues.

 

Please join us for these informative talks, and consider staying for the afternoon class (separate event), taught by Ron Cody. For more information on the afternoon training, please visit www.basug.org/events/t200912.htm

 

Immediately following the meeting, we will provide an informal light buffet lunch for all morning meeting and afternoon training attendees. We hope you can stay for this opportunity to network and socialize with your fellow SAS users.

 

Please note that we are in a new venue for this meeting!

 

Topic:

Coders’ Corner 2009

 

When:

Wednesday, December 2, 2009, 8:15am – Noon

 

Where:

 

Microsoft New England Research & Development Center

One Memorial Drive

Conference Center First Floor NEW LOCATION!

Cambridge, MA 02142

(857) 453-6000 

Directions:

Please visit: http://www.basug.org/directionsNERD.html

 

How:

Individual, On-Line Registration Required. No Email!

 

To register:

Please visit: http://www.basug.org/register.php3

 

Contact:

If you have questions about the meeting contact Lori Goldman (Goldman@hcp.med.harvard.edu) or Sally Skinner (sally.skinner@tufts.edu)

 

 

Agenda

 

8:15

Sign in and Refreshments

 

8:45

Announcements

 

9:00

Acquiring Data via FTP in SAS

by James Zeitler, Harvard Business School

 

9:12

Array Processing with the Generic Macro

by Victor Pontes, Inatec, Inc.

 

9:34

Statistical Significance Testing with Stratified Samples

by Joan Yu, Data Rewards

 

9:44

Break

 

10:00

Forest Plots in SAS

By Meena Doshi, New England Research Institutes

 

10:12

Hash Table Lookup from SAS Data Files

by Sue Szubzda, Ingenix

 

10:24

Creating Dynamic HTML pages with PC SAS, as a prelude to SAS INTRNET or Integration

by Qiang Hou, Harvard Pilgrim Health Care

 

10:41

The Colorful ODS

by Ali Sabouri, Blue Cross Blue Shield of Massachusetts

 

10:51

Break

 

11:11

Using Proc Transpose to Flip Your Output

by James K. West PhD, MA Department of Public Health

 

11:23

Using Systask to Dynamically create Libnames

by Ray Cloutier, Health Dialog

 

11:35

Combining Data Sets Containing Character Variables of Different Lengths

by Ron Cody

 

11:50

Closing remarks & announcements

 

12:00-1:00

Informal buffet lunch (provided by BASUG)

 

1:00-4:30

Afternoon Training – separate eventclick here for details: www.basug.org/events/t200912.htm

 

 

*Note: Times (and sequence) are approximate and subject to change. Please re-visit the BASUG website ( www.basug.org) for updated information.

 

 

Abstracts and Bios

 

 

Acquiring Data via FTP in SAS

by James Zeitler, Harvard Business School

 

A simple FILENAME filref "pathname"; statement serves the purpose when reading from and writing to simple text files on disk. But SAS accommodates more exotic file access methods that open up new vistas to the data step. We'll look at one of these, the FTP access method, and see how it facilitated the acquisition of multiple data files from an external source.

 

James Zeitler is a Senior Information Research Specialist at Harvard Business School, assisting HBS faculty with research across a variety of business disciplines, and has recently passed the quarter-century mark as a SAS user.

 

Array Processing with the Generic Macro

by Victor Pontes, Inatec, Inc.

 

The SAS Macro facility is a powerful tool for automated code generation. Most programmers regularly use Macro programming to add efficiency to Base SAS code generation. But the macro language isn't just for generating Base SAS code; it is used to simplify the production of repetitive Macro code as well. We will explore some of these possibilities using the Generic Macro, which will simplify the production of macro code by taking the code itself as a parameter.

 

Victor Pontes has spent his entire professional life programming, mostly using SAS but also as an applications programmer. Five years ago he founded Inatec, Inc., through which he works as an independent consultant, primarily in the areas of pharmaceutical and biomedical research. He is enamored with the SAS macro facility and enjoys using it creatively and testing its limits.

 

Statistical Significance Testing with Stratified Samples

by Joan Yu, Data Rewards

 

When stratified sampling is used to select marketing treatment and control groups, one group, either the treatment or the control, needs to be weighted for the statistical comparison analysis. This talk will show how to use Proc Surveymeans to calculate the correct standard errors for statistical significance testing and to compare the results with those from Proc Ttest using a weight statement. The talk will be given in the context of marketing, and the analysis approach can be applied to other fields as well.

 

Joan has focused on the use of data and technology to improve marketing performance throughout her 12-year career after she received her Ph. D. and M.S. degrees in statistics from Northwestern University. Her experience extends across a number of industries including retail, financial services, pharmaceutical and high tech. At Data Rewards, Joan is responsible for delivering analytic solutions that are actionable and cost effective.

 

Forest Plots in SAS

by Meena Doshi, New England Research Institutes

 

There is a notion that SAS/GRAPH is one of the most difficult components of SAS software; however once understood it is a very powerful tool for creating a wide range of charts, maps and plots in a variety of patterns, colors and arrangements. Simple SAS/GRAPH options will be demonstrated in easy steps to show even the beginner programmer or graph user how to make a customized forest plot in SAS.

 

Meena Doshi is a Statistician at the New England Research Institutes, Inc in Watertown, MA. She has been using SAS for seven years, both in the fields of Marketing Analysis and Public Health.

 

Hash Table Lookup from SAS Data Files

by Sue Szubzda, Ingenix

 

Accurate data is key in the health care industry. One aspect to improving the health care industry is providing the most accurate medical history for patients possible.  By grouping all health care records of the patient together, we can provide a more accurate assessment of the member's health care history and improve case management. 

This presentation demonstrates the use of hash table lookup methodology to update the member id on historical medical records with the most recent member ID; thereby grouping all historical claims of the member.   

Likewise, grouping all of a provider's patient(s) records under one provider ID provides a more accurate assessment of the provider's performance.  This demonstration also shows how the process is used simultaneously to update the provider's id on historical medical records with the most recent provider id. 

Sue Szubzda has worked at Ingenix (Waltham, MA) for 4 years as an Applications Developer on the Custom Client Development team writing/testing production code for client specific requests.  Previous experience includes contracting with Hewlett Packard providing support for Sales and Marketing and 18 years in the electric utility industry (Southern Company/Alabama Power) primarily using SAS to develop customer rate plans and to convert/test a consolidated billing system (nine billing systems combined into one).  She holds undergraduate degrees in Business Education from the University of Alabama and in Computer Science from the University of Alabama at Birmingham.

 

 

Creating Dynamic HTML pages with PC SAS, as a prelude to SAS INTRNET or Integration

by Qiang Hou, Harvard Pilgrim Health Care

 

This paper introduces a simple technique to create SAS dynamic web applications, employing PC SAS and freely-available open-source software, as an alternative to the SAS products IntrNet and Integration Technologies. Such an easily-implemented alternative may be especially desirable in the early or prototype stages of design or software development.

 

Utilizing PC SAS together with freely-available open-source software (Apache and PHP), this paper shows how to develop interactive web pages that can process a user request from any web browser, and convey results back to the user in real time, displaying virtually any type of data. Using an example of an interactive web page that charts the closing price of a user-supplied stock from a user-supplied date to a current date, the paper explains in detail how the technique works.

 

Qiang Hou works for Harvard Pilgrim Health Care as a business analyst performing data analysis and predictive modeling. He holds a doctorate in Statistics from Tufts University. Prior joining Harvard Pilgrim he worked for Hypnion (which became part of Eli Lilly in 2007) as a biostatistician. His recent interest is on healthcare data mining and predictive modeling.

 

The Colorful ODS

by Ali Sabouri, Blue Cross Blue Shield of Massachusetts

 

Ever since its inception in SAS® V8, Output Delivery System (ODS) has been a powerful tool in presenting reports in various formats (i.e. HTML, RTF, PDF, etc.). Examples here will show you how to efficiently apply simple codes to make your presentations more colorful. You’ll learn how to use Proc Report in conjunction with ODS to “traffic light” cell values within your HTML output.

 

Ali Sabouri is a Senior Network Technical Consultant at Blue Cross Blue Shield of Massachusetts in Boston where he does data analyses, programming and reporting. Ali is a SAS Certified Programmer and has been using SAS® since 1993 mostly within healthcare industry.

 

 

Using Proc Transpose to Flip Your Output

by James K. West PhD, MA Department of Public Health

 

I often expend my programming energy and expertise coaxing numbers from data sets of thousands of records without paying any attention to the lay out of my output. I think this is proper. If I’d wanted to be a graphics artist, I wouldn’t have taught myself SAS, right? Wrong. I find that my output, though technically successful, doesn’t present the findings meaningfully -- there are columns where I want rows and rows where I want columns. Proc Transpose is a handy tool to flip these files.

 

In my work, I often have a large number of data sets with individual records like births, deaths, and hospitalizations. I aggregate the values of certain fields on individual records to state totals by age, sex, and race/ethnicity (strata). Here is my method using PROC TRANSPOSE.

 

James K. West has been a demographer and epidemiologist at the Massachusetts State Department of Public Health for eight years. Before working at the Department, he earned his PhD in Public Policy at the University of Massachusetts Boston, where he received one of the last degrees of the 20th Century on December 31, 1999 at age 49. For most of his 30s and early 40s, he worked in the computer field in the Boston area at jobs ranging from software support at Honeywell to network analyst at Bolt, Beranek and Newman. He is a self-taught SAS programmer and avid reader of postcolonial fiction.

 

 

Using Systask to Dynamically Create Libnames

By Ray Cloutier, Health Dialog

 

Spending a little extra time planning your directory structure can have big returns. When projects are stored in similar folder structures, it becomes very easy to create cross-project applications. This paper will demonstrate how to use the systask command to dynamically create libname statements to perform cross-project analysis.

 

Ray Cloutier (pronounced Clue-Chi) works at Health Dialog as a Analytics Manager for the Clinical Development department. He has been using SAS for over 10 years and currently uses SAS to manipulate over 200 terra bytes of medical claims data.

 

Combining Data Sets Containing Character Variables of Different Lengths

By Ron Cody

 

This SAS tip has two purposes: First, it solves a fairly common problem that is encountered when you want to combine two data sets and, second, it uses an interesting (and powerful) technique where you have a SAS program that writes a SAS program that then gets submitted. If you want to append values in one data set to the end of another data set, you can either use PROC APPEND, or use a SET statement that lists the names of the data sets you want to combine.

 

However, what if the two data sets have the same variable names, but some of the character variables have different lengths in the two data sets? Furthermore, the character variables stored with longer lengths are not consistent in one or the other data set. You cannot use PROC APPEND because that procedure will use all the attributes from the base data set, which will truncate any character values that are longer in the second data set. You can use a SET statement, but you will have to manually enter the length of each character variable (in a LENGTH statement) from either data set one or data set two, whichever is longer.

 

Because this is a fairly common problem, there is a macro that combines two data sets and automatically uses the maximum length for each character variable.

 

Ron Cody is a retired professor from the Robert Wood Johnson Medical School who now works as a private consultant and a national instructor for SAS Institute. A SAS user since 1977, Ron's extensive knowledge and innovative style have made him a popular presenter at local, regional, and national SAS conferences. He has authored or co-authored numerous books, such as Learning SAS by Example: A Programmer's Guide; Applied Statistics and the SAS Programming Language, Fifth Edition; The SAS Workbook; Cody's Data Cleaning Techniques Using SAS Software; Longitudinal Data and SAS: A Programmer's Guide; and SAS Functions by Example, as well as countless articles in medical and scientific journals.

 

 

BASUG Membership

 

Keep your BASUG Membership up-to-date! Print out a membership form from our website, and bring it with you to the meeting (along with your check). Please note that we have discontinued Corporate Memberships – we now provide only Individual Membership at $30 annually. For more information on our membership policy, or to print out a form visit:

http://www.basug.org/basugj.shtml

 

 

Directions to Microsoft New England Research & Development Center

 

Please visit: http://www.basug.org/directionsNERD.html

 

 

BASUG Contacts

 

Mailing Address:

 

BASUG

PO Box 170253

Boston, MA 02117

 

To email our Webmaster: basugwm@basug.org